| As the demands of precision and intelligence in the manufacturing are getting higher and higher,the urge of detection methods of diamond wheels trimming and detection is becoming stronger.Visual detection system can detect the trimming of diamond wheels through a non-contactable and real-time way.Though theoretical errors of visual system and preliminary detection results of grinding wheel outline can be calculated in present researches,comparative researches of the precision of the visual detection system are rather rare and there is still improvement room for visual detection system in efficiency and intelligence.This thesis chooses the visual detection system of diamond wheels as the research object.In hardware components,it mainly discusses the detection precision of arc and straight line features.It builds a full set of detection system of diamond wheels trimming.After the trimming and detection of grinding wheels are finished,the detection of image measuring instrument and three coordinate detection of arc and straight line features are conducted.This thesis analyzes and evaluates present visual detection systems by comparing different kinds of high-precision detection methods.In software improvement,the efficiency of the visual detection system is improved and the craftsmanship database is preliminarily designed.(1)Based on the laser composite intelligent trimming wheel equipment,the diamond wheel trimming detection system is built.The system integrates mechanical method,electro sparking and laser trimming.It is composed of CCD camera,clamping illuminating instrument,detection software,etc.The detection basis is grinding double mapping which can collect photos,reserve test information and output detection results.(2)Duplication plates' grinding duplication tests are conducted by diamond wheels against arc and straight line features.For arc features,arc radius is ground and the designing radius is 5 mm.For straight line features,the straightness of duplication plates is ground to reflect the grinding results.In the test,the order is graphite platesoft steel plate-graphite plate.The rotating speed of the spindle and feed rate are strictly controlled to advance duplication plates' shape quality.(3)Comparative tests and data analysis of shaped grinding wheels outline are conducted.Multilevel detection tests are conducted for arc and straight line features.Arc features detect the radius and straight line features detect straightness.Through visual detection system,image measuring instrument and three coordinate detection,pre-grinding graphite plate,soft steel plate,post-grinding graphite plate and grinding wheels are measured.By comparing statistics,it is found that the detection results of straightness is quite excellent,using soft steel plate as duplication plate can reflect the precision even better,but grinding wheels are worn.(4)To meet the requirements of automation and intelligence of visual detection system,the software of visual detection system is improved.While meeting the detection precision,the thesis adopts double three B spline interpolation algorithm to improve the efficiency of template matching algorithm in detection system.The thesis chooses pictures in duplication as verification image,the efficiency is improved by 9.6times compared to previous algorithm,and the precision change is less than 0.4%.The craftsmanship database is added.While the function of adding,deleting,changing and checking is fulfilled,the relationship of each entity in visual detection system is analyzed,the E-R picture is drawn and the logic layer design is finished. |
PDF Full Text Request